Extended range simulations with the regional ETA model to diagnose model performance

The implementation of changes into atmospheric models implies the need of effective verifications that determine the quality of the model predictions. A methodology to diagnose the deficiencies of the regional model is to run extended range simulations. In this study assessment of the Eta model quality is carried out by performing one-month continuous integrations over South America. The regional Eta model runs operationally at Centro de Previsão de Tempo e Estudos Climáticos (CPTEC) providing short term forecasts over the continent. The resolution is 40 km in horizontal and 38 layers. Two land-surface scheme are tested in the model, one with bucket scheme and the other with the CAPS scheme. Simulations are done for summer and winter months. Profiles of bias and standard deviation of humidity, temperature and zonal and meridional wind averaged over different areas of the integration domain are shown. Diagnostic of temporal evolution of domain averaged variables: temperature, surface pressure, specific humidity, precipitation and kinetic energy show the stability of the model in both versions. These diagnostic sugest the usefulness of the model for future regional climate simulations.